Stable and fast network connectivity is critical to cloud computing. Aspects, such as high-availability, throughput, and redundancy are core to its functionality. Still, there are many more additional measures necessary to deliver a reliable but secure service. In this post, as a key member of our networking team I’m going to outline four key steps we’ve taken to give customers the best networking experience we can. When we talk about customers having a great networking experience that means not only high availability and reliability but also high throughput and low latency from anywhere in the world.
Here are the four key steps our team has taken to deliver the networking capabilities to our customers:
Blended Tier-1 Carriers
In order to achieve a high level of redundancy and failover CloudSigma always hosts itself in carrier-neutral data centers. This allows us to blend together multiple fibre optic networks from major internet connectivity providers. Thus, we can create a profile for our customers that is better than anyone connectivity company can provide.
We’ve also made efforts to ensure that we can failover automatically and seamlessly between providers should one network be experiencing difficulties or if one or more of our own routers suffers an outage. By maintaining these connections backed up 24/7, CloudSigma guarantees 100% network uptime. We are able to deliver a fast connection to and from customer virtual instances. Last but not least, our customers benefit from the great speed that these lines. We never throttle customer transfer rates. Combined with our unique software driving our cloud, customers can stream at multiple gigabit speeds without issues.
Another initiative undertaken by our networking team is our efforts to peer with the major IXPs throughout the world. Some of those are DECIX, Swiss-IX etc. This allows us to hand traffic and to receive traffic directly from to and from the end networks. This ensures faster and more reliable routing of traffic which means lower latency for customers. As a result, CloudSigma customers benefit from excellent visibility over internet routes as well as even more redundancy across networks. Above all, this means less downtime for customers, generally lower latency and faster load times.
Connected with the GÉANT fibre network
CloudSigma is connected to the GÉANT academic network, reaching a research and academic community with over 50 million users at 10,000 institutions across Europe, as well as interconnections with Europe’s National Research and Education Networks (NRENs). The CloudSigma cloud network is available to route traffic directly over these high capacity, low latency networks for the global research community. Combined with the growing public data sets we host for people such as the European Space Agency. It means scientists and students alike can get access to the best data faster and more reliably than ever before.
As a public cloud with thousands of customers, it is inevitable that we will be the subject of denial of service and distributed denial of service attacks (DDoS). In recent years these have become almost daily events for anyone providing services with network connectivity. Our networking team has had more than a few late nights dealing with incidents in order to protect customer connectivity. It’s that experience that we bring to bear in dealing with the threats as they morph and develop over time.
All our clouds are protected by our own custom anti-DDoS system. It gives me and the rest of the networking team to tweak and adjust how we protect our cloud locations over time. It means that for the vast majority of incidents, customers never even are aware that there was a networking event. We are able to mitigate attacks usually within seconds.
Additionally for customers wanting extra peace of mind and protection, we recently partnered with Black Lotus to offer a great way to protect any IP addresses in CloudSigma that a customer might want custom protection over.
I hope this post gave you some insights into the various aspects of networking that we concentrate on when it comes to the plumbing that sits behind the CloudSigma cloud.